tdd-workflows-tdd-red
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user data via the
$ARGUMENTSvariable and interpolates it directly into the prompt template used by thetest-automatorsubagent. This allows a potentially malicious user to provide input that overrides the subagent's instructions. - Ingestion points: Found in the 'Prompt Template' section and at the end of the
SKILL.mdfile where$ARGUMENTSis used. - Boundary markers: Absent. There are no delimiters (e.g., XML tags, triple quotes) or specific instructions provided to the agent to treat the
$ARGUMENTScontent strictly as data rather than instructions. - Capability inventory: The skill invokes the
Tasktool with theunit-testing::test-automatorsubagent to generate and document tests. - Sanitization: No validation or sanitization of the input arguments is implemented in the skill's instructions.
Audit Metadata